Restoration Opportunity Module¶

File: computing/misc/restoration_opportunity.py

Overview¶

This module generates forest landscape restoration opportunity layers based on WRI (World Resources Institute) data. It produces both raster and vector outputs with restoration potential classifications.

Purpose¶

Processes restoration opportunity data for: - Forest restoration planning - Land degradation assessment - Conservation prioritization - Distributes to GEE, GeoServer, and STAC

Architecture¶

flowchart TD
    A[Celery Task Triggered] --> B[ee_initialize]
    B --> C[Load MWS ROI]
    C --> D[clip_raster]
    D --> E[Load WRI Restoration Raster]
    E --> F[Clip to ROI]
    F --> G[Export Raster to GEE]
    G --> H[Sync to GCS]
    H --> I[Sync to GeoServer]
    I --> J[Generate Raster STAC]
    J --> K[generate_vector]
    K --> L[Calculate Area by Class]
    L --> M[Export Vector to GEE]
    M --> N[sync_layer_to_geoserver]
    N --> O[Return True]

Components¶

Main Task: generate_restoration_opportunity()¶

Location: Line 25

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | state | str | State name | | district | str | District name | | block | str | Block/tehsil name | | gee_account_id | int | GEE account identifier |

Returns: bool - Whether layer was successfully synced to GeoServer

Helper Functions¶

clip_raster()¶

Location: Line 55

Clips the restoration opportunity raster to ROI and exports to GEE.

generate_vector()¶

Location: Line 114

Converts raster classes to vector features with area calculations.

Processing Logic¶

1. Raster Processing¶

Data Source:

restoration_raster = ee.Image(
    GEE_DATASET_PATH + "/WRI/LandscapeRestorationOpportunities"
)

Classification Values: | Value | Label | Description | |-------|-------|-------------| | 0 | Excluded Areas | Not suitable for restoration | | 1 | Mosaic Restoration | Mixed land use restoration | | 2 | Wide-scale Restoration | Large-scale forest restoration | | 3 | Protection | Areas needing protection |

2. Vector Generation¶

For each class, calculates pixel area within each MWS:

flowchart LR
    A[Raster Class] --> B[Create Mask]
    B --> C[Pixel Area Image]
    C --> D[reduceRegions]
    D --> E[Area per MWS]
    E --> F[Convert to Hectares]

Processing:

for arg in args:
    mask = raster.eq(ee.Number(arg["value"]))
    pixel_area = ee.Image.pixelArea()
    forest_area = pixel_area.updateMask(mask)

    fc = forest_area.reduceRegions(
        collection=fc, 
        reducer=ee.Reducer.sum(), 
        scale=10, 
        crs=raster.projection()
    )

    # Convert to hectares
    value = ee.Number(value).multiply(0.0001)
    feature = feature.set(arg["label"], value)

Integration Points¶

computing/misc/restoration_opportunity.py
├── computing.utils
│   ├── sync_layer_to_geoserver() # GeoServer sync
│   ├── save_layer_info_to_db()   # Database persistence
│   └── update_layer_sync_status() # Status tracking
├── utilities.gee_utils
│   ├── ee_initialize()           # GEE authentication
│   ├── check_task_status()       # Task monitoring
│   ├── valid_gee_text()          # Text sanitization
│   ├── get_gee_asset_path()      # Asset path generation
│   ├── is_gee_asset_exists()     # Asset existence check
│   ├── sync_raster_to_gcs()      # GCS synchronization
│   ├── sync_raster_gcs_to_geoserver() # GeoServer sync
│   ├── export_raster_asset_to_gee() # Raster export
│   ├── export_vector_asset_to_gee() # Vector export
│   └── make_asset_public()       # ACL management
├── utilities.constants
│   └── GEE_DATASET_PATH          # Dataset path
└── computing.STAC_specs
    ├── generate_raster_stac()    # Raster STAC generation
    └── generate_vector_stac()    # Vector STAC generation

Output¶

Raster Output¶

Dataset Name: Restoration Raster

STAC Layer Name: wri_restoration_raster

Vector Output¶

Dataset Name: Restoration Vector

Feature Properties¶

Each MWS feature contains:

{
  "uid": "unique_mws_id",
  "Excluded Areas": 123.45,
  "Mosaic Restoration": 456.78,
  "Wide-scale Restoration": 789.01,
  "Protection": 234.56
}

All area values are in hectares.

Usage¶

from computing.misc.restoration_opportunity import generate_restoration_opportunity

result = generate_restoration_opportunity.delay(
    state="Rajasthan",
    district="Jaipur",
    block="Sanganer",
    gee_account_id=1
)

Pattern Classification¶

This module follows the Raster-to-Vector Pattern:

flowchart TD
    subgraph Pattern
        A[Load Raster] --> B[Clip to ROI]
        B --> C[Export Raster]
        C --> D[Sync Raster to GeoServer]
        D --> E[Calculate Class Areas]
        E --> F[Export Vector]
        F --> G[Sync Vector to GeoServer]
    end

This pattern is shared with: - stream_order.py

Dependencies¶

• ee (Google Earth Engine Python API)
• Celery - Distributed task queue